A friend asked me how many Republican presidents have carried the electoral votes of New York State. It’s a simple enough question. But I think it may be masking something else. My friend may be wondering if the present is monolithic: that the way things are is the way they have always been, and always will be.

New York is currently overwhelmingly Democratic—so much so that it’s nearly inconceivable that New York would vote Republican for president. In 2004, 58.37% of the vote in New York was for John Kerry, the Democrat (he earned 48.3% of the popular vote nationwide). In 2008, Barack Obama took 62.88% of New York (52.9% nationwide). In 2012, Obama got 63.35% of New York (51.1% nationwide). In 2016, Hilary Clinton took 59.01% of New York (48.2% nationwide). And in 2020, Joe Biden garnered 60.87% of New York (51.3% nationwide).

But reality is not quite so unchanging.

There have been 42 elections with a Republican candidate on the ballot for president (the first was John C. Fremont in 1856). In those elections, New York has voted for the Republican candidate 20 times, and 22 times for the Democrat. When New York votes Republican, that candidate has won 17 times (New York voted for unsuccessful Republicans Fremont in 1856, Hughes in 1916, and Dewey in 1948). When New York votes Democratic, that candidate has won 15 times (New York voted for unsuccessful Democrats Seymour in 1868, Tilden in 1876, Humphrey in 1968, Dukakis in 1988, Gore in 2000, Kerry in 2004, and Hilary Clinton in 2016).

So yes, in the current era, New York state voting Republican is highly unlikely. But such was not always the case. And it may not always be the case in the future. The key take-away from these numbers is: things change. The Republicans may be able to divorce themselves from the insanity of the Trumpian party, and once again field rational candidates for national office. The Democrats may become unhinged through their own acceptance of a radical fringe. The one thing we know for certain is that humans are very good at extrapolating trends, but absolutely terrible and foreseeing inflection points in those trends.